The most common mistake people make in mid-career transition to AI

AI will impact many jobs and COVID has only accelerated this trend

Based on my teaching at the University of Oxford for AI, we see a big demand for people wanting to transition to AI due to the current environment.

If you are at the early stage of your career (less than two years out of Uni), a transition to AI is relatively easier

But you may need a different strategy for a mid career transition to AI

Why?

In the early stage of your career, all things being equal, you would add more value than the company pays you

But about 4 plus years into your career, that changes

At that time, you are moving away from core development (directly creating value) and also your are being paid more - leading to a decline in value in pure economic terms

(I know this is a generalization - it does not apply to everyone but it is valid overall)

You then end up with a variety of titles that give incremental pay basically do the same thing

At this point, many a career could languish

Hence, it is at this point that people explore alternate careers paths especially to AI

But mid-career transition to AI is a bit different

The biggest mistake by far which people make - is ignoring what they already know

The general thinking goes like this

You focus on learning Python and equate knowledge of coding with AI

Coding is indeed necessary but is not a sufficient in itself

Because ultimately to add value, you would need to more than coding alone

So, now let's consider a different starting point

You learn Python but you also consider what you know

To give you an example, suppose you are an engineer

An engineer could model a problem using specific techniques such as mathematical modelling (see my previous blog on this topic HERE)

A recent paper expands this idea by integrating Physics-Based Modeling With Machine Learning

Once you model a problem, you could work on model evaluation criteria for specific problems and compare them to industry benchmarks

In other words, you could co-relate problems in your industry to data science

The good thing is, data science will be in every field

Hence, the trick to transition to data science is to find and solve big problems in your industry based on ideas you already know from your experience in the industry

Image source: General Assembly - United Nations

Tags: dsc_ai, dsc_tagged

2 members like this

Comment

You need to be a member of Data Science Central to add comments!

Join Data Science Central

Comment by Jörg Narr on November 26, 2020 at 1:03am: Thank you, Ajit. Fully agree. Times will change. Senior roles can bridge the gap between Data Scientists and Business/Use Cases.

Comment by ajit jaokar on November 25, 2020 at 11:24pm: thanks for your comments @Jorg

I wanted to stress two points in this post

a) Midcareer transition is not the same as early stage due to the need to add more value

b) The previous blog was an idea for illustration

c) re " I do see many Data Scientists working on problems and optimizing solutions in companies which remind me more on "investigating" than solving issues." this is indeed true. However, I see that times will change. ie people will go from investigation to solving more complex problems
thanks for your comments! rgds ajit

Comment by Jörg Narr on November 25, 2020 at 10:30pm: I generally like the post but I think particularly the conclusion is quite brief and seems to me to be mainly used to create more traffic on previous posts. Nevertheless I second your thoughts. In my experience sufficient AI knowledge, frameworks, and technology are here whilst clever ideas to apply AI to real-world problems is quite rare. I do see many Data Scientists working on problems and optimizing solutions in companies which remind me more on "investigating" than solving issues.

The most common mistake people make in mid-career transition to AI

You need to be a member of Data Science Central to add comments!

TechTarget Articles